Trump issues AI, data center position; Anthropic CEO defends record on safety, sales to China
Artificial superintelligence companies are facing heavy scrutiny amid the midterm elections as Congress leaders weigh regulatory action, public safety checks and moral and ethical reviews of their cutting-edge work.
Pentagon lays out plan for autonomous warfare command by 2027
Leaders of the Pentagon’s Project Agincourt outlined a roadmap to establish an Autonomous Warfare Command by Oct. 1, 2027, aiming to speed the development and fielding of drones and other autonomous systems.
In an open letter to military and industry partners, Owen West and Senior Chief Max Strasiser said the command would pair robotics companies directly with frontline units and judge new systems by their performance in exercises and combat — not paperwork milestones.
The effort responds to the battlefield impact of low-cost drones and seeks to help U.S. forces adapt faster, the letter said.
“Urgency does not excuse imprecision,” West and Strasiser wrote, pledging to evaluate which technologies should scale and cut those that do not meet operational needs.
Tensions flare over city council vote on data centers: 'We can't keep trusting your word'
Fox News correspondent Madison Scarpino joins "America’s Newsroom" to report on the Memphis City Council rejecting an 18-month pause on new data center development amid resident concerns.
Tensions flare in Tennessee as the Memphis City Council rejects an 18-month pause on new data center development despite pushback from local residents.
Nvidia’s Huang touts Europe’s AI ambitions, takes the stage with robot ducks
Nvidia CEO Jensen Huang praised Europe’s engineering base as a potential advantage in artificial intelligence and robotics, announcing expanded partnerships with European manufacturers and startups during the company’s GTC conference in Berlin on Wednesday.
“Every company in Europe will become an AI company,” Huang said.
Europe, he added, has unmatched expertise in building precise mechanical and electronic systems — skills he said could help the region become a producer of AI technology, not just a consumer.
Huang shared the stage with small, AI-trained duck-shaped robots developed by French startup Pollen Robotics with Hugging Face. The robots performed somersaults, though some struggled when Huang told them to take a nap.
One later faced the wrong way as he finished his speech.
“Pay attention, stay focused,” he told it.
Nvidia said it has committed nearly $10 billion to Europe’s technology ecosystem. Germany’s Agile Robots plans to produce 1,000 to 2,000 small humanoid robots next year in collaboration with Nvidia, according to CEO Zhaopeng Chen.
Production is expected to begin late this year.
Nvidia also showcased a new design for factory-built data center modules using its Vera Rubin platform, saying the facilities could be deployed in as little as two weeks, down from six months.
What is PHP? The behind-the-scenes code powering much of the web
PHP is a programming language that powers websites behind the scenes, assembling pages on a server before they appear in a visitor’s browser.
It runs on nearly 70% of websites whose programming language can be identified, according to web technology tracker W3Techs.
PHP also powers WordPress, the platform used by about 40% of all websites, including blogs, business pages, government portals, online shops and news sites.
Because PHP works behind the scenes, most website visitors never interact with it directly. The software is free and open source, maintained by a community of developers.
Publisher drops award-winning author after finding AI markers in novel, says he denied using tools
Canadian publisher Les Éditions du Boréal said Wednesday it had ended its professional relationship with Haitian-Canadian author Thelyson Orelien after finding evidence that artificial intelligence was used in his award-winning debut novel.
“C’était ça ou mourir” (“It Was Either That or Die”) won several literary awards and was slated for translation into multiple languages before a little-known account on X alleged last month that it had been written almost entirely with AI.
Orelien previously denied using AI but acknowledged instances of past plagiarism.
Reuters was not immediately able to reach him for comment.
Boréal said its investigation included an external analysis that found traces of AI markers in the metadata of certain documents. The publisher said it could not establish how AI was used in the book’s creation.
“We still cannot know in what capacity AI was used in the creation of ‘It was either that or die,’ and undoubtedly we will never have a definitive answer,” Boréal wrote in a statement.
“After the allegations, we continue to trust Thelyson Orelien, who assured us he had never used such tools. This assertion has now been proven false.”
Boréal said it would not reprint the novel or publish future works by Orelien.
Reuters contributed to this report.
Anthropic taps former Harvard science dean for AI oversight body
Anthropic said Wednesday that Christopher Stubbs, former dean of science at Harvard University, has joined the company’s Long-Term Benefit Trust, an oversight body with the power to appoint a majority of the AI startup’s board.
The trust advises Anthropic’s leadership on major decisions involving AI risks and societal impacts. Its members include former Federal Reserve Chair Ben Bernanke and national security expert Richard Fontaine.
Stubbs is an experimental physicist and Harvard professor of physics and astronomy. He was part of the High-Z Supernova Search Team, which helped discover that the universe’s expansion is accelerating.
Anthropic, developer of the Claude family of AI models, operates as a Public Benefit Corporation, a structure that allows it to balance shareholder interests with broader societal considerations.
FBI, NSA urge organizations to patch critical PHP flaws or take systems offline
The FBI and NSA are warning organizations to urgently address two critical vulnerabilities in PHP that could let attackers remotely run code and potentially compromise sensitive data and networks.
The FBI, in an X post, urged organizations to update affected PHP installations immediately, following guidance from PHP’s developers. If an update cannot be made right away, the FBI recommends taking the affected system offline until it is fully protected.
The flaws affect PHP (Hypertext Preprocessor), programming language used to build and run websites. PHP developers have released a fix, while cybersecurity agencies have urged site operators to update or temporarily take vulnerable websites offline.
A web server can use PHP to assemble a page’s content and send it to a visitor’s browser. It powers widely used platforms such as WordPress, along with many other websites and online services.
The NSA issued a similar warning, saying the vulnerabilities could provide a path to a server’s file system or database. It advised organizations using PHP to be prepared to install an update and said it would provide more specific guidance.
"Effective Immediately, all organizations using PHP for their systems are advised to be prepared for an update addressing two critical vulnerabilities. PHP is widely used for web applications. These vulnerabilities often provide a direct path to the server's file system or database. If an immediate update is not feasible, it is recommend that organizations take the system offline until it can be fully mitigated. NSA will follow up shortly with more specific guidance."
The FBI also urged organizations to inventory where PHP is running, prioritizing internet-facing systems and products that include PHP, and to monitor web servers for suspicious activity. Signs of possible exploitation could include PHP processes unexpectedly launching command shells or downloading files.
Organizations that find signs of a compromise should isolate affected systems, preserve web server and PHP logs, and report the incident to the FBI or the Internet Crime Complaint Center, the agency said.
OpenAI says frontier AI found PHP vulnerabilities before attackers
OpenAI said its frontier AI models discovered critical PHP vulnerabilities before attackers did, adding a new dimension to a security scramble that has already prompted government agencies to take websites offline.
The flaws affect PHP (Hypertext Preprocessor), programming language used to build and run websites. PHP developers have released a fix, while cybersecurity agencies have urged site operators to update or temporarily take vulnerable websites offline.
A web server can use PHP to assemble a page’s content and send it to a visitor’s browser. It powers widely used platforms such as WordPress, along with many other websites and online services.
Security experts have warned that AI could help attackers exploit such flaws at scale.
Dutch and Finnish government services were among those taken offline as a precaution.
Reuters contributed to this report.
AI could accelerate PHP flaw attacks as government websites go offline
Government websites in Europe and elsewhere went offline Wednesday as agencies scrambled to address critical vulnerabilities in PHP, a widely used programming language.
PHP developers released a fix and urged users to update.
Cybersecurity agencies in Europe, Australia, Japan and Britain advised operators to patch their systems or take vulnerable sites offline. Dutch and Finnish government services were among those disrupted, while the Netherlands’ cybersecurity agency also took its website offline as a precaution.
Security experts warned attackers could scan for vulnerable servers at scale.
AI may speed up that process, said Lukasz Olejnik, a visiting senior research fellow at King’s College London: “With AI exploitation may happen instantly, on a large scale.”
Reuters contributed to this report.
GlobalFoundries chief says Japan-backed chip plants will build US ‘reserve capacity’
GlobalFoundries Executive Chairman Tom Caulfield said a planned investment of up to $21 billion from Japan will help build two semiconductor plants in upstate New York, which the company will operate.
The partnership with the U.S. and Japanese governments will create “reserve capacity” so chip supply can move ahead of demand and help avoid shortages like those seen in 2021, Caulfield told FOX Business’ Lydia Hu on "The Claman Countdown."
Caulfield also defended the U.S. government’s equity stake in GlobalFoundries, saying investments in critical industries can bolster national and supply-chain security while generating returns.
AI uncovers ‘most severe vulnerabilities ever discovered,’ cybersecurity firm says
Cybersecurity company Wiz said its AI tools uncovered a flaw in the PHP programming language, describing the vulnerabilities as “the most severe” ever discovered by AI.
Wiz is a cloud security company that helps organizations find and fix security risks across their cloud environments.
Super Micro stock jumps 11% as forecast signals resilient AI demand
Super Micro Computer shares surged Wednesday after the AI server maker forecast quarterly revenue and profit margins well above Wall Street expectations, signaling continued demand for data center equipment.
The stock rose 11.6% to $36.25, its highest level in a week. Super Micro expects first-quarter revenue of $18.5 billion to $19.5 billion, compared with analysts’ average estimate of $15.23 billion, and projected margins of 12% to 12.5%.
Dell shares gained 5.1%, while Hewlett Packard Enterprise rose 3.2%. Major technology companies are expected to spend more than $700 billion on AI infrastructure this year, though investors remain focused on whether those investments will deliver lasting returns.
Reuters contributed to this report.
AI stocks rise as SpaceX and Super Micro results lift Wall Street
Wall Street opened higher Wednesday, with AI stocks gaining after quarterly results from SpaceX and Super Micro Computer boosted investor appetite.
The Dow Jones Industrial Average rose 0.30% at the open, while the S&P 500 gained 0.31% and the Nasdaq Composite climbed 0.67%. Netflix shares slipped after its forecast disappointed investors.
Reuters contributed to this report.
AT&T CEO says customers will control whether AI agents can manage their accounts
AT&T customers will decide whether AI agents can manage their accounts, CEO John Stankey said, responding to Meta’s new Muse Spark AI model.
Any account transactions made by an AI agent would still need to pass the company’s validation and authentication checks, Stankey said.
Reuters contributed to this report.
Trump demands data centers make their own power, pay community ‘dividends’
President Donald Trump said early Wednesday morning that large data centers should generate their own electricity and provide financial benefits to the communities where they are built.
“New and large, in some cases gigantic, Data Centers are mandated by me to make their own electricity,” Trump wrote in a Truth Social post.
"As the President of the United States, I have been watching Data Centers very closely, and have come to a very strong conclusion. New and large, in some cases gigantic, Data Centers are mandated by me to make their own electricity, and give or sell any excess back to the local community, or grid, and therefore there is electric price increase. In fact, there could be a substantial decrease. Other elements likewise! If companies want to make their Data Centers popular, pay the community, and/or individuals within the community, a 'dividend.' I have watched the process closely, and Data Centers that were highly unpopular in a certain area, have become extremely popular when a financial deal is struck. MAKE DEALS, and get them built!!! GREAT FOR THE U.S.A. President DJT"
He said facilities should sell or return excess power to local communities or the electric grid, potentially lowering electricity prices.
Trump also urged technology companies to offer residents a “dividend” to build support for data center projects, which have faced opposition over their substantial power and water demands.
“MAKE DEALS, and get them built!!! GREAT FOR THE U.S.A.,” Trump wrote.
The president said the administration’s policy is intended to prevent consumers from absorbing the cost of powering the AI industry.
The comments follow similar proposals from Anthropic and Elon Musk to distribute some of the financial gains from artificial intelligence to the public.
OpenAI president backs AI safeguards as political group he funded fights regulation
OpenAI President Greg Brockman has publicly pushed for AI safeduards while a political group backed by a $25 million donation from Brockman and his wife campaigns against regulations that could “unduly constrain or delay AI development.”
Leading the Future is spending millions in at least 48 House and Senate races, backing candidates who support limiting state-level AI laws and creating a single federal standard.
AI safety advocates say the political spending raises questions about whether OpenAI executives genuinely support rules that could slow model releases or reduce industry profits. OpenAI said the donation was made in Brockman’s personal capacity.
Reuters contributed to this report.
AI regulation fight pits national standards against state rules
A political group funded by OpenAI President Greg Brockman and major venture capital investors is backing candidates who support a federal AI standard that could override state regulations.
Leading the Future and an affiliated nonprofit have promoted federal pre-emption, a policy that would prevent states from passing or enforcing their own AI rules.
The group is spending millions in congressional races, including contests involving candidates who oppose “heavy-handed” regulation.
OpenAI rival Anthropic has taken the opposite position, giving at least $40 million to a political action committee opposed to federal pre-emption.
Anthropic CEO Dario Amodei has called the proposal “too blunt.”
Reuters contributed to this report.
OpenAI executive’s political giving sparks AI regulation questions
A $25 million donation from OpenAI President Greg Brockman and his wife to a pro-AI political group has drawn scrutiny as OpenAI executives call for federal safety safeguards.
The group, Leading the Future, opposes regulations that “unduly constrain or delay AI development” and is targeting at least 48 congressional races.
Critics accuse tech executives of supporting AI safety rhetoric while backing candidates opposed to restrictions on the industry.
“There are leaders in industry who want to pay lip service to regulation,” said Charlie Blaettler of the Guardrails Alliance, a group formed to oppose Leading the Future.
Reuters contributed to this report.
El-Sayed backs govt ownership of half of AI industry as Rogers vows to block ‘big government’ agenda
Democratic Senate candidate Abdul El-Sayed called for sweeping government control of the artificial intelligence industry, including public ownership of roughly half its value, drawing a sharp rejection from Republican rival Mike Rogers.
“And I think the government should hold about 50% of the value,” El-Sayed said during a Detroit Economic Club event, arguing that AI companies should be rebuilt as public-interest corporations.
El-Sayed said the proposal would help protect workers from AI-driven job losses and direct more wealth toward schools, child care and other public services.
He also called for stronger regulation of large technology companies, arguing that a handful of corporations have become too dominant to face meaningful competition.
Rogers rejected El-Sayed’s approach, warning that government ownership, higher taxes and additional regulations would raise costs for Michigan families and discourage investment.
“I’ll be damned if I’m going to let Detroit become a Democratic socialist city,” Rogers said. “They’re talking about big government, government control, more rules, more regulation, another Green New Deal.”
Rogers said Michigan should focus on lowering utility and auto insurance costs, expanding manufacturing and rebuilding its defense industrial base.
“We need to open up a sign and say, Michigan is open for business again,” Rogers said. “Let’s be the work belt of the Midwest again.”
Likely voters give Democrats an edge on handling artificial intelligence in new poll
Likely voters trust Democrats over Republicans to handle artificial intelligence by 11 percentage points, with 48% choosing Democrats and 37% choosing Republicans, according to The New York Times/Siena poll released early Wednesday morning.
The results come as AI companies face mounting scrutiny over safety, regulation, foreign sales and the technology’s impact on jobs, national security and the economy.
Democrats’ advantage on artificial intelligence has grown since September, the Times noted.
The poll did not specify which AI policies or candidates drove voters’ preferences.
The survey of 3,494 likely voters was conducted Oct. 12-19, two weeks before the midterm elections.
Anthropic CEO touts China sales restrictions in fiery defense of AI safety record
Anthropic CEO Dario Amodei defended the company’s AI safety record Tuesday while highlighting its decision to restrict sales of AI services to Chinese state-controlled companies.
“We were the first, and still only, frontier company to restrict the sale of AI services to PRC-controlled companies,” Amodei wrote in a nearly 2,000-word post on X overnight, using the Chinese name for its country (People's Republic of China).
Amodei was responding to Nathan Calvin, chief strategy officer for AI-safety nonprofit Encode, who accused Anthropic of undermining its safety commitments through more than $500 billion in computing deals, increased reliance on Claude to develop new models and changes to its safety policy.
The Anthropic CEO countered by listing nine decisions he said were commercially costly but advanced safety, including limiting access to the company’s Mythos model, dedicating 5% of Anthropic to a public-benefit fund and using stringent safeguards against biological-weapons misuse.
Amodei acknowledged that Anthropic had made mistakes, including how it communicated a draft federal AI safety bill. The proposal would allow the government to seek a court order halting a model posing an imminent risk of more than 50 deaths or serious injuries, or more than $1 billion in damage.
He said Anthropic remains committed to proposing options for regulating the pace of AI research by Jan. 15, 2027, or explaining why it will not do so.
Live Coverage begins here
Covered by: Eric Mack